Brendan Bohannan
I argue that microbiome research needs mathematical models rather than loose metaphors, and suggest a productive modelling strategy. I focus on a quantitative genetics framework that explores the differential contributions of host genes, microorganisms and environment to system-level traits (of the host and its microbes). This raises conceptual questions about whether host-associated microorganisms are genes or environments, and how to think about heritability and fitness in such systems. Of particular philosophical interest are the implications for evolutionary modelling of mixed-species entities, and how far a quantitative genetics framework can be extended across the tree of life to multispecies consortia.
